Hi all,
I have another mistery to solve. I am trying to infect primary hyppocampal neurons with lentiviruses expressing my protein and then iresGFP. construct work good and are green when transfected. I produced my lentivirus and used to infect 293T cells to obtain titer and the 293T cells are green after infection. When I infect my neurons however, they are not green, I tryied to perform immunofluorescence and western blot and in both cases I see my protein overexpressed but no GFP.
Can anybody help me?
Gilda
Hello Gilda,
The problem is the IRES. Indeed, in some experimental context you can have one of the two transgenes which is not expressed and it is often in vivo contexts. I have had the same problem few years ago : I validated my construction in vitro in transfection and transduction but after stereotactic injection (mice) I lost the expression of one transgene.
I suggest you to use another strategy involving a 2A sequence, in this case, you don't have this problem, the two transgenes are expressed in comparable level in any experimental context
http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0018556
I can help you to design the construction if you need
Nicolas
Hi Nicolas,
thank you very much for your answer. I can't believe these things can happen even though I have seen them happening many times now...science is weird too many times!!! It is very strange as one of my colleague was using this lentivirus before me and she was having green neurons...maybe now they decided they had enough...
I will give a look at that paper but the problem is that we do not have the backbone constructs...
thank you so much anyway you were extremely helpful
Gilda
Effectively, this differential of work is strange. In my case, my vectors work poorly but it was reproductible.. it is the same lentiviral production of your colleague or a new one? It is the same brain region? When you have a good GFP expression, do you have a good expression of your other transgene?
In all cases, if you look the bibliography in vectorology, you can observe that 2A constructions are more reliable that IRES constructions in all the experimental contexts tested.
Nicolas
Ires is much weaker than a well designed Kozak driving your gene. Also, the expression level (of everything) is much higher during packaging, when you have multiple copies of free floating plasmids. I've seen GFP drop close to, or below the level of detection after infection.
I had similar problems also with T2A (Kozak included). I was lucky as I could use ubiquitous promoters. I solved the issue by generating bidirectional dual promoter lenti backbone (gene of interest driven by promoter in one direction and fluorescent protein driven by another promoter in the opposite direction). Works very well in vitro and in vivo (brain) and tested with several genes. I also generated unidirectional dual promoter lenti backbones that work well.
Nicolas is 100% right. LV IRES-GFP constructs drive me crazy, out of the ten or so I've made this year I can see GFP fine after transfection but almost no GFP after infection. The IRES results in expression of GFP that's just too low to detect. You can probably see it OK in 293Ts because they're a lot easier to infect than neurons. Have you considered using AAV instead? If your gene isn't too big to package then AAV-DJ works really well in neurons.
- Badges
- Science method